JPS63188283A - Character reader - Google Patents

Abstract

PURPOSE:To perform the reading of a large number of and a various kinds of slips with high efficiency, by performing a character recognition processing after storing a quantized image on which reading control (FC) information is attached in a large capacity data storage device. CONSTITUTION:In a magnetic disk device 14, the FC information and the quantized image are stored as a sheet of slip image. A control part 11 takes out the quantized image at every slip unit from the magnetic disk device 14, and sends it to a character recognition processing part 16, and at the character recognition processing part 16, the FC information attached on the quantized image is read out first, and after information required for control being obtained, the quantized image is taken out, and the image of the slip is produced in a two-dimensional memory space. And by segmenting the character image of one character from the slip image, the character recognition processing is performed. In such a way, it is possible to perform the reading processing of the next slip without waiting the completion of the reading processing of one sheet of slip, and to improve working efficiency remarkably.

Description

[Detailed Description of the Invention] [Object of the Invention] CJl Business Application Field) The present invention relates to a character reading device that reads characters written on a form, and particularly relates to a character reading device that reads characters written on a form, and particularly to a character reading device that reads quantized images of a form, such as a magnetic disk device. The present invention relates to a character reading device configured to perform character recognition processing after being stored in a data storage device such as a data storage device.

(Prior technology) Character reading device! (hereinafter referred to as OCR), optically scans an input document while conveying it, quantizes this scanning signal to obtain a quantized image, and cuts out a character image from this quantized image. It is configured to perform character recognition processing.

Here, the above-mentioned OCR is configured to read characters written on a form based on reading control information (hereinafter referred to as FC information) determined by the type of form. Reading requires FC information corresponding to various forms. Furthermore, this FC
The information is information for reading control given to OCR, such as sheet size, reading lines, number of fields,
Consists of data such as mask start, character pitch, font, subset, etc.

However, in conventional OCR, the process from obtaining a quantized image of a form to character recognition processing is performed in real time, so the next form cannot be read until the reading process for one page is completed. For example, when reading a large number of forms, it is necessary for the operator to intervene until the reading process for all the forms is completed, which reduces the operator's work efficiency.

Furthermore, in conventional OCR, when reading a wide variety of forms, it is not possible to set the FC information for the next form until the reading process for one form is completed, resulting in extremely poor reading efficiency. There were drawbacks.

(Problems to be Solved by the Invention) As mentioned above, conventional systems have been configured to read characters written on forms in real time, which has the disadvantage that the operator's work efficiency decreases when reading a large number of forms. This method has the disadvantage that the reading efficiency in reading various types of forms is poor.

The present invention has been made in view of the above points, and an object of the present invention is to provide a character reading device that can efficiently read a large number of various types of forms without reducing the work efficiency of an operator.

[Structure of the Invention] (Means for Solving the Problems) That is, the character reading device according to the present invention is a character reading device that reads characters written on a form based on reading control information determined depending on the type of the form. In the device,
The reading control information of various forms is stored in the reading control information storage means in advance, and the predetermined reading control information is applied to the quantized image of the form obtained by optical scanning in the reading II (
The control information is read out from the control information storage means and added, or designation information that designates the predetermined read control information is added.

Then, the quantized image added with the read control information or the quantized image added with the specified information is stored in a quantized image storage means consisting of a data storage device in Daeyong-eup, such as a magnetic disk device. A quantized image is retrieved for each form unit from the mother-child image storage means, and character recognition processing is performed based on the reading control information added to the retrieved quantized image, or character recognition processing is performed based on the added specified information. The reading control information is configured to be read from the reading control information storage means and subjected to character recognition processing.

゛ (Function) In the character reading device configured in this way, even if the reading process for one document is not completed, the quantized images of a large number of documents can be stored in advance in, for example, a magnetic disk device. You can leave it there.

In this case, predetermined read control information or designation information specifying this read control information is added to the maternalization image stored in the magnetic disk device.

Therefore, after storing the quantized images of a large number of forms in the magnetic disk device, if the quantized images are retrieved for each form from this magnetic disk device, the quantized images can be read based on the read control information added to the retrieved quantized images. It is now possible to perform character recognition processing by reading out the reading control information corresponding to the specified information that has been added from the reading control information storage means. Also in reading, efficient reading processing can be performed without reducing the work efficiency of the operator.

(Example) Hereinafter, a character reading device of the present invention will be explained with reference to the drawings. FIG. 1 is a block diagram showing the circuit configuration of a character reading device according to an embodiment.
3, a magnetic disk device 14, a magnetic disk device 15, a character recognition processing section 16, an interface section 17, a keyboard (hereinafter referred to as KB) 18, and a CRT display (hereinafter referred to as CRT) 19 are connected, respectively.

The scanning unit 13 has a function of photoelectrically converting an input form using an image sensor such as a COD scanner, and outputting a quantized image. The magnetic disk device 14 is
The quantized image obtained by this scanning unit 13 is stored.

The magnetic disk device 15 stores FC information of various forms in advance. The character YLm processing unit 16 has a function of cutting out a character image from a quantized image and subjecting it to character recognition processing.

The interface unit 17 has a function of transmitting and receiving data to and from the outside.

The operation of a character reading device having such a configuration when reading a large number of various types of forms, for example, will be described below. In other words, a magnetic disk device in advance! 15 stores FC information for various forms, and the operator
Predetermined FC information can be specified by operating the KB 18 without displaying the menu on the RT 19. Therefore, by classifying a large number of documents to be read by type, the operator can specify FC information in batches.

By specifying such FC information, a large number of forms classified by type are sequentially fed to the scanning unit 13. The scanning unit 13 obtains quantized data for one document by scanning the input document with an image sensor such as a COD scanner, and outputs this quantized data to the magnetic disk device 14 .

The magnetic disk device 14 sequentially stores the quantized data output from the scanning section 13. In this case, as described above, the FC information for each batch is specified by the operator, and the control unit 11 retrieves the specified FC information from the magnetic disk device 15.
Information is added to the quantized data.

In this way, as shown in FIG. 2, the magnetic disk device 14 stores FC information (FC) and quantized images (IM).
will be stored as one sheet of form data. This storage of quantized images is performed for all the forms fed to the scanning unit 13, and the predetermined FC information and quantized images are stored in pairs in the magnetic disk drive@14 in the order of classification by type. The generated form data is stored.

When the quantized images of all the documents to be read are stored in the magnetic disk device 14, the control unit 11 retrieves the hanging images for each document from the magnetic disk device 14, and processes the retrieved quantized images for character recognition. Send to Department 16. The character V&identification processing unit 1B first reads out the FC information added to this quantized image to read the size and shape of the form, coordinate position of input data, character size, character type, etc. necessary for reading control. After obtaining the necessary information, the quantized image is extracted and an image of the form is created in a two-dimensional memory space. Then, character aX processing is performed by cutting out a character image for one character from this form image based on the FC information.

The recognition results obtained by the character recognition processing unit 16 are edited according to the FC information and stored in, for example, the magnetic disk unit W115. Therefore, the reading results of all the forms are stored in the magnetic disk device 1115, and the reading results are sent to the outside, such as the host device, via the interface unit 17.

In this manner, the character reading device of this embodiment is configured to perform character recognition processing after the quantized image of the form to be read is stored in the magnetic disk device of Daeyongsan. Therefore, when reading a large number of documents, the reading process for one document is not completed.
The next form can be read. For this reason,
It is no longer necessary for the operator to intervene until the reading process of all the documents is completed, as in the past, and the work efficiency of the operator can be greatly improved. In other words, operator intervention is only necessary up to the preprocessing of storing the quantized image, and the postprocessing character recognition process can be performed unattended.

Furthermore, in the case of this embodiment, a predetermined FC is applied to the quantized image.
Because information is added and stored on a magnetic disk device,
There is no need to set the FC information for the next form after the reading process for one form is completed as in the past, and the efficiency of reading various types of forms is improved.

In this embodiment, predetermined FC information is added to the quantized image and the magnetic disk i! [14] However, the present invention is not limited to this; for example, the third
As shown in the figure, it is also possible to add specification information (FC number) specifying FC information and store it in the magnetic disk device [i14]. In this case, the character recognition processing section 16 reads predetermined FC information from the magnetic disk device 15 from the above specified information, and then performs character recognition processing based on the read FC information.

In this way, when specifying information specifying predetermined FC information is added to the quantized data, the amount of data on one sheet is reduced, and the capacity of the magnetic disk device 14 is reduced compared to the case of the above embodiment. can do.

Further, in the above embodiment, the scanning unit 13 is configured to scan with one scanner, but it is also possible to obtain quantized images of multiple forms at the same time by using multiple scanners, for example. .

[Effects of the Invention] As described above, according to the present invention, using a large-capacity data storage device such as a magnetic disk device, predetermined FC information or designation information specifying predetermined FC information is transmitted to this device. Since the system is configured to perform character recognition processing after storing the added quantized image, it is possible to efficiently read a large number of various types of forms without reducing the operator's work efficiency.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a circuit configuration of a character reading device according to an embodiment of the present invention, FIG. 2 is a diagram showing a form data storage state of a magnetic disk device according to the same embodiment, and FIG. FIG. 3 is a diagram illustrating the storage state of form data in the magnetic disk device according to the embodiment. DESCRIPTION OF SYMBOLS 11... Control part, 13... Scanning part, 14... Magnetic disk device, 15... Magnetic disk i! Stomach, 16.
...Character recognition processing section, 17...Interface section. Applicant's Representative Patent Attorney Takehiko Suzue Figure 1 Figure 2 Figure 3

Claims (1)

[Scope of Claims] A character reading device that reads characters written on the form based on reading control information determined depending on the type of form, comprising a reading control information storage means that stores reading control information of various forms in advance; , a form reading means for reading a form by optical scanning and outputting a quantized image; and predetermined reading control information is read from the reading control information storage means and added to the quantized image obtained by the form reading means. , or information adding means for adding specification information specifying the predetermined reading control information, and quantized image storage means for storing a quantized image to which the reading control information or specification information has been added by the information adding means. , A quantized image is extracted for each form from this quantized image storage means, and character recognition processing is performed based on the reading control information added to the extracted quantized image, or character recognition processing is performed on the attached specified information. A character reading device comprising character recognition processing means for reading out corresponding reading control information from the reading control information storage means and performing character recognition processing.